Bacteria are generally considered to be a single cellular, free living organism, yet they carry toxin or suicidal genes on their genome. These toxins are usually coexpressed with their cognate antitoxins. The sets of these toxin and antitoxin pairs are most often encoded from the single operon, and termed TA systems. The first such example was found in plasmids and bacteriophage, and required for post-segregational killing' (15, 17). The TA systems are now widely found not only on plasmids and on bacteriophages, but also in the chromosome of almost all bacteria. E. coli carries at least 29 sets of toxin-antitoxin (TA) systems, but the mechanisms of action are deciphered in only a small number of them. Known toxins inhibit a wide range of cellular functions including DNA, mRNA, 30S and 50S ribosome subunits and cell division (17, 30, 37, 38). The exact role of these toxins on bacterial physiology is not yet clearly understood. One possible role is to regulate cell growth under certain severe growth conditions for survival (bacteriostatic). Another theory is that toxins are suicidal, killing unwanted cells in order to maintain a desired population (bacteriocidal), as seen in post-segregational killing (6). The ymgG-ymgD and dicB-ydfD operons from the E. coli chromosome have been identified as possible TA systems using a computer program to predict TA systems (RASTA) in bacterial genomes (27). Both ymgD and ydfD encode for a short peptide (consisting of 109 and 65 amino acid residues, respectively) which are toxic to cell upon expression. However, their cellular targets have not been identified. On the basis of preliminary results from my study, the targets of these toxins appear to be the cell wall, causing a rapid decrease in cell viability upon their induction. This is the first example of TA toxins targeting cell walls to cause cell death. Interestingly, YmgD is produced with a signal peptide and is the first toxin discovered to be secreted into the periplasmic space. In this application, I will attempt to decipher the exact molecular reactions in peptidoglycan (PG) biogenesis, which these toxins target. I will propose a research project examining how YdfD reduces the degree of PG crosslinks. When YdfD expression is induced, 4'-3' crosslinks (between 4' D-alanine residue of one PG stem and 3' meso-DAP of another) are specifically reduced. I predict that YdfD may inhibit the formation of 4'-3' crosslinks, thus weakening the crosslinking of PG polymer causing cell to lyse. In order to unveil the mechanisms of how YdfD reduces the crosslinking, I will first examine the effect of YdfD expression on the specific enzymatic reactions necessary for polymer formation using purified proteins in in vitro assay systems. Chromatographic analysis of precursors will be also performed to examine the accumulation/lack of biosynthetic intermediates. The peptidoglycan purified after the induction of YmgD lacks most of the peptidoglycan crosslinked lipoprotein. Combining this result with our preliminary result that YmgD toxicity was reduced in the lipoprotein deficient strain, we hypothesized that YmgD toxicity is lipoprotein dependent. We have found that YmgD interact with two subunits of the outer membrane lipoprotein translocator, Lol complex. Lol complex is responsible for translocating outer membrane lipoproteins, including Lpp, from inner the membrane to the outer membrane. There are more than 90 outer membrane lipoproteins that are known and serve essential functions. Inhibition of Lol systems is known to cause cell lysis due to the accumulation of lipoproteins in the inner membrane. YmgD may exert its toxicity through inhibition of Lol system, causing improper localization of lipoproteins. I will test this hypothesis through purification of lipoproteins from membranes by differential centrifugation and look for any unusual accumulation of mislocalized lipoproteins. The effect of YmgD on the activity of Lol complex will also be assessed using in vitro system. The research proposed here will advance the understanding of complex bacterial physiology controlled by various TA systems. Since a number of pathogenic bacteria also carry numerous TA systems, and are potentially involved in pathogenicity, further characterization of the TA systems will provide a clue to develop a new strategy for treating pathogens. Also, since the toxins target the cell wall and cause rapid cell lysis, our investigation will directly provide us a new means to examine the basic biology of cell wall biosynthesis, and will reveal a new target of drug design and also lead to develop novel antibiotics.